Quantitative Methods for Psychologists I

1

• Quantitative Methods for Psychologists I
• QM4?1
• for PhDs in Psych (so you know some stats and
  psych)
• www.fiu.edu/dwright/qm4psych

My slides are all on the web at least the day
before the lecture, so don't write lots of stuff.
(at least you can find parking at 830 ... after
the first week)
2
Topics today

• Psychology meets statistics
• Outline, what I think we will do this term
• The structure of lectures and the sign up!
• You wanna be a millionaire?
• Briefly, what is SPSS/PASW
• What is R and how to get it
• - Reading data into R
• Today's journal assignment

3
Scientists (including psychologists) rely on
statisticians - every psychology article is
filled with stats - without statistics, we
would lack a (good) language to describe
our empirical results - we would lack a
convenient way to make our data appear
convincing/credible information Interactions
between psychologists and statisticians can be
difficult.
4
Statistics are, for psychologists

• For discovering patterns in data/nature
• turning data into information
• A method for communicating this information
• to tell a story with the data/information

5
Is it in SPSS?
Best estimators (and showing it) Mathematics that
won't help our audience understand our
papers Things we don't (can't) understand
6
(No Transcript)
7
Computing and Statistics
From version 18 called PASW. Now owned by IBM.
and hundreds more
8
Should people fear stats be doing it?
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
Topics for the term
14

• Boxplots, quantiles, histogram  intro to
  probability.
• Means, standard deviation, skewness,
  transformations, making bad graphs.
• Categorical data, bar charts, measurement,
  ranking
• Sampling and allocation
• Inference, confidence intervals, power
• t-test, Wilcoxon, NHST and alternatives
• ANOVA intro
• Regression intro
• Chi-square
• Advanced glance (if time)
• Presentations x2 or x3

15
Assessments

• Essay 25 due on the final presentation
• Final presentation 15
• Journal/portfolio 50
• How-to pages 10

16
Structure of typical class

• Dan gives a background lecture with some "how to"
• One (or more) of you describes how you ran the
  procedure with SPSS, the difficulties you faced,
  and what you learned
• One (or more) of you describes how you ran the
  procedure with R, the difficulties you faced, and
  what you learned
• Do some analyses on the computers
• Discuss portfolios from the previous week

17
Sign up list

• You must sign up for at least one SPSS topic.
• You must sign up for at least one R topic.
• You cannot sign up for the extra slots until all
  slots are filled.

18
Groups of

• From different programs.
• You have 1 million for research.
• What research would you do?
• Think of all the costs.

19
Did you ...

• Find out each other's names and where each is
  from?
• Did you remember your salary, health care,
  pension, room costs, etc.?
• Did you make a time table?
• Do you have dissemination plans?

20
(No Transcript)
21
What is SPSS/PASW?
SPSS Statistical Package for the Social
Sciences PASW Predictive Analytics SoftWare

• Widely used in academia (and increasing outside
  of academia).
• Back in the 60s it was free (peace, love, etc.),
  but became a commercial product (i.e., robbed
  from the proletariat), and now owed by IBM.
• http//www.gnu.org/software/pspp/

"Data control and IBM,Science is mankind's
brother" Jefferson Airplane
22
Who has used SPSS before?
23
Pretty user friendly package
24
R Statistical environment and language
What is this?
25
R is freeR is object oriented (everything is an
object)
26
http//cran.r-project.org/
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
Reading in Data

• Typing at the command line
• Reading from text/dat file
• Reading from SPSS
• Reading from Excel

We will describe making data using probability
functions
31

32
chile lt- read.table("http//www.sagepub.co.uk/wri
ghtandlondon/chile.dat",headerTRUE)chile lt-
read.table("c//temp/chile.dat",headerTRUE)chile
lt- read.table("d//chile.dat",headerTRUE)
33
Reading from SPSS(and Minitab, SAS, Stata,
SYSTAT, etc.)

• Need the package foreign (R core members et al.,
  2009)
• Steps needed to access package
• First time on a machine
• install.package("foreign")
• choose a mirror
• Afterwards
• library(foreign)
• Downloads lots of functions for reading and
  writing from other statistics packages

34
Can't do from a web page
Double slashes needed here and for some other
commands
Get warnings often
35
(No Transcript)
36
Excel

• Excel stores information in odd way
• Packages xlsReadWrite (Suter et al., 2006) and
  RODBC (Lapsley Ripley, 2008)
• (e.g., read.xls function)
• I find these awkward the official advice is to
  save to text file from Excel.
• Alternative Go into Excel and Ctrl-A, then
  Ctrl-C
• (i.e., select all, and copy)
• chile lt- read.delim("clipboard")

37
(No Transcript)
38
Return to the Chile
39
hist(LENGTH, xlab"Length in cm")
40
Study Groups

• Makes statistics easier. You learn more provided
  you are an active member of the group.
• Good if across labs/silos.
• I'll send an email with everyone's email.

41
See me afterwards if ...

• You have taken graduate statistics already
• If you have no statistics
• If you are not doing a PhD in psychology
• If you are signed up for next week's
  presentations
• And anything else that seems appropriate

42
This week's portfolio

• Download R onto a computer (preferably your own)
• http//www.uk.sagepub.com/upm-data/26922_Wright_C
  hapter_One.pdf
• http//journal.r-project.org/2009-1/RJournal_2009
  -1_Chambers.pdf
• Make a histogram
• Find or make data
• Play with options (use book and help facility)
• Include graph (right click on graph to save),
  commands, and output in a Word file (copy into
  Word).
• Keep this as entry 1 of your portfolio. Bring
  portfolio to class each week.